At low single-client volume it’s roughly a wash with going AI Ark-first. The edge isn’t that Prospeo is cheap — it’s that QuickEnrich is a flat $400/mo tool that finds ~70% of emails for effectively free, and the same AI Ark budget stretches across ~4× the clients (used only as fallback, not as the primary engine).
| Approach | What the ~$2,449/mo plan covers | $ / send-ready lead |
|---|---|---|
| AI Ark only (old) | Finds contact and email, all-in | $0.017 |
| Prospeo flow (new) | Prospeo sources; cheaper tools find the email | $0.013 |
| Tool | Role | Unit cost |
|---|---|---|
| Prospeo | Sourcing (contacts, no email) | $0.0114 / credit · 25 contacts/cr |
| AI Ark | Source+email bundled / fallback finder | $0.002449 / credit (1M plan) |
| QuickEnrich | Primary email finder | $400/mo FLAT (~$0 per email) |
| Website crawl | Backup email finder | ~free static / Firecrawl render |
| Reoon | Email verification | free (lifetime) |
| The job | Cheaper option | Cost |
|---|---|---|
| Find the contact only (no email) | Prospeo (~2.7×) | $0.0005 vs $0.0012 |
| Find contact + email, bundled | AI Ark | $0.0025 vs $0.0035 |
| Email found by flat QE / free crawl (~70%) | Prospeo flow | ~$0 marginal |
| Email found by AI Ark fallback (~30%) | AI Ark on its own | we pay twice |
| Blended (real-world) | Prospeo flow, ~20% | only via flat QE |
Prospeo can’t reveal emails itself, so on the ~30% QuickEnrich misses we pay Prospeo to source and AI Ark to find — on those, AI-Ark-first would’ve been cheaper. The flat QuickEnrich subsidy on the other ~70% is what tips the blend.
| Scenario | Contacts to source | $ / month | $ / lead |
|---|---|---|---|
| AI Ark-first | — | ~$2,060 | $0.0171 |
| Prospeo flow — Microsoft parked (today) | ~1.54M | ~$1,600 | $0.0133 |
| Prospeo flow — Microsoft unlocked | ~730K | ~$830 | $0.0069 |
~43–55% of valid leads sit on Microsoft email systems we can’t send to yet — already paid for, just parked. Unlocking Microsoft sending roughly halves cost per usable lead (we stop sourcing 2× the volume to discard the Microsoft half). It hits both sourcing approaches equally, so it doesn’t change which source is cheaper — it lowers the floor for both.
Two distinct models. Prospeo is contact-first (find named people by ICP, then their email). GMaps is business-first (scrape local venues in kiosk cities, email attached).
GMaps gets the email for free with the scrape ($0.005 all-in incl. email) but only matches businesses in kiosk cities. Prospeo sources any ICP contact cheaply but must find the email separately — which is why the email waterfall (and the flat QuickEnrich subsidy) is where the Prospeo cost model is won or lost.
| Lever | Effect |
|---|---|
| 1. Unlock Microsoft sending | Frees ~half of already-paid leads · ~halves cost/usable lead · biggest win |
| 2. Add cheap pay-per-find tools (LeadMagic / Blitz) | Catch QE-misses before the AI Ark fallback · 0-on-miss = low risk · shrinks priciest line |
| 3. Expand geo / ICP | Market exhaustion (recency) is the real volume ceiling, not cost |
Lead cost is already ~a penny per usable lead. The binding constraints at scale are send capacity, deliverability, and addressable inventory — not the price of leads. Unit economics improve as the roster grows (flat fees amortize).